PH热榜 | 2026-01-20

Congrats to the Team 🎉🎉🎉

daily.dev sits on my new tab. Good for the morning scroll. Hiring on top makes sense. Just don’t let it turn into LinkedIn vibes. Keep it opt-in and quiet and I’m game. Open source is a nice touch. I’ll try it with a niche role and see how it goes.

I've been trusting daily.dev with being my #1 aggregator of dev news, and rather trust them instead of LinkedIn or other platforms with finding interesting work. Since MoBerries is gone, there hasn't been a single recruiting platform that gives me as a developer the feeling of being valued. Thank you daily.dev.

Been using daily dev since 2022! - This is super exciting to see

Smart move turning an existing developer community into a hiring channel. Feels way more natural than traditional job boards.

Congrats on this 6th launch!!!

Recruiter is brilliant! Usual channels (LinkedIn, generic job boards) are so noisy and full of uninterested applicants, being able to tap into a community of active, passionate devs who are already engaged every day feels like a much smarter way to find great fits ahah

How does the matching work on the recruiter side? Is it skill-based, activity-based, or a mix? And do devs opt-in explicitly for recruitment outreach?

This is a great initiative. Just came across after seeing the launch.

Keeping it genuine for developers will make it a great community and authentic recruitment place

Hey there, long time Daily.dev reader (if you aren't reading it already, you should be!) I am super excited to see someone disrupting tech hiring!

Huge day for the team! 🎈

Recruiting in tech often feels broken, so we're opening up our network to make those connections more authentic.
Would love to hear what you all think about the video and the product. Do you think this changes the 'passive candidate' game?

In daily.dev our focus was always creating the best space for developers, with recruiter, this has not changed, it expands the things developers can do and accomplish without ever leaving daily.dev, the place they feel most comfortable.

Bootstrapping and building daily.dev Recruiter was a great ride with many technical but also product challenges and don't make me start talking about creating support to parse any JD in existence 😂🤞

We are just starting but looking forward to seeing what everyone thinks!

Your Product page and website looks great and all of your messaging seems very well thought out.

How long has it taken you to craft your messaging and did you do a lot of AB testing to ultimately arrive at the conclusions you did?

A daily.dev Recruiter is a breath of fresh air in the recruiter world. No more spam, no more optimizing for quantity. Ask us anything on your mind

As someone who’s spent years designing developer products, this launch was less about “adding recruiting” and more about removing everything broken in how hiring feels today.

Most recruiting tools are designed for volume. More profiles, more messages, more dashboards. We designed daily.dev Recruiter around the opposite constraint: trust.

From a UX perspective, the hardest part wasn’t matching algorithms. It was designing a system where developers never feel hunted, spammed, or tricked into conversations they didn’t ask for.

Double opt-in. High context. No cold access.
If it doesn’t feel respectful, we don’t ship it.

Curious to hear how this resonates with folks who’ve felt the pain on either side of hiring.

Been a huge fan of @daily.dev and @nimrodkramer for years and was one of their first subscribers! I have no doubt this will be just as awesome! Keep up the good work!

I really want this to turn out well. I have been a long-time daily dev reader; it has been part of my daily source of dev information anytime I open my PC, so it would also be cool to see job opportunities here as well

Congrats on the launch!
Do you match the best potential roles to developers? If not, how do you make sure this doesn't become another place for recruiters to spam?

@nimrodkramer Double opt-in intros from “real behavior” is a strong wedge; at scale the hard part is turning 1B+ sessions into trustworthy intent signals without gaming + while keeping developer trust/privacy airtight.

Best-practice: event-stream (Kafka/PubSub) → feature store, real-time aggregates in ClickHouse, plus an explicit consent/audit ledger + per-company rate limits so “warm” stays warm.

Q: how are you modeling/validating “intent” (rules vs embeddings/ML), and do developers get transparent controls to view/revoke what signals drive matches?

Daily Dev Recruiter is incredibly unobtrusive and intuitive, speaking from a developer standpoint. As a member of the daily dev community, I visit the site every day as it's my homepage and my "new tab" page on my browser. When I first saw a job recommended to me, it matched my interests, level, and skill sets. My reason for not applying at that moment was due to having recently starting a new project. However, I was impressed with the accuracy and how vetted the recommendation was both from the recruiting side and the candidate side. It'll be my go-to in the future.

congrats on the launch! Does the app only attracts SWE? what about other jobs/tasks like tutor/design etc.? Will you cover other areas later?

Hey Product Hunt 👋🏻

I’m Suyash, one of the folks building Interactpitch. Really excited to share this today.

Fundraising is deeply personal. You spend weeks refining your story, your numbers, and your vision. Then you send a pitch and wait, hoping it lands the way you intended.

What is InteractPitch?

InteractPitch is an interactive pitch deck experience that helps founders tell their story more clearly and understand how investors engage with it before you get on the first call.

Instead of passively flipping through slides, investors can explore the pitch at their own pace, ask questions as they come up, and go deeper where they’re curious. An avatar of yourself guides the experience, helping investors move through the pitch the way you would in person.

The problem

Most fundraising conversations start with uncertainty.

After sharing a pitch, founders have no visibility into how it was experienced. You do not know what resonated, what caused confusion, or what investors discussed internally. When the call begins, you are often reacting instead of leading.

The solution

InteractPitch brings real insights before the actual pitch happens.

Founders can see how investors engaged with the pitch, which sections held attention, and what questions came up along the way. This makes the first call more focused, more confident, and far more productive.

Features

-Interactive pitch avatar

-Mobile-first pitch experience

-Text + voice conversations

-Slide-aware answers

-Custom knowledge grounding

-Engagement insights

Why it’s different

Most tools help you send a pitch. Some help you track it. InteractPitch helps you understand how it was experienced.

InteractPitch is not a replacement for live pitching, it is an enhancement to the pitch deck experience. It makes pitches easier to explore, easier to understand, and easier to discuss, while giving founders the context they need to show up prepared.

Who it’s for

Built for founders raising capital and teams running important investor conversations.
Useful for pitch decks and pre-meeting reviews.

We have put a lot of care into making this feel human, thoughtful, and genuinely helpful.
Would love your feedback and thoughts.

Thanks for checking out InteractPitch 💫

Where is the InteractPitch pitch deck for InteractPitch? Would be the perfect demo!

As a former entrepreneur who scrambled to raise funds, this is a very personal and valuable product for me.

Very interesting idea, team!

Hi @suyash_karn @utsav_interactai @sudhir_p product is really good, tried and tested it loved the first version. Here are my 2 cents on the product:

1. Import Feature is amazing - I have tried multiple online tools but this has outperform import feature especially pixel by pixel
2. Personalization is all eveyone needs - give user feature to make avatar from their image - this is doable - I have seen many opensource projects - let me know if u need help in this.
3. Manage Access needs to be easier and better UI/UX flow wise
4. Chat reply needs to be improved - Tried asking question related to slide but slightly in different way - output was basic - not upto the mark. When question is outside the deck then chat fails by mentioning ots outside deck
5. Use better contextual model or better give user power to add FAQs or set of questions with related answers - this will be the game changer post product usage and i beleive this can be game changer
6. When someone ask in chat - avatar stops and mess start from here
7. I also noticed some context mess - bot is replying to previous questions instead of currently askled questions - this happens only when i send somethign in chat while bot is speaking - as per my experiment

Overall Liked the first version fo your product, its great. This has good potential for startups and B2B usage as well.

Thanks!

Congrats on the launch! Looks like a great product for agencies as well. How do you train AI avatars to provide the right answers?

that is what founders need!! does the avatar strictly follow a script, or can it dynamically answer basic investor queries on the spot?

Not gonna lie, I’ve always wondered what happens after I send my deck into the void 😂 Love the idea of turning it into a guided experience. Trying it now.

Seeing real investor questions about my deck is a huge upgrade from just knowing their time spent on each slide. Definitely trying this out.

Congrats team!

Nice. Are there predefined templates and greater image generation and charting tools?

@suyash_karn Kudos , seems interesting, i tired it

though seems like PDF is not a right option to start with creating a Interact Pitch(message: "Failed to convert PPTX slides to images"})

Looking forward to trying soon

I have got into a lot of VC conversations. Everytime the first thing they ask is that do you have a deck. I send it and then it's gone into their analyst data. I have no understand as to what are the next steps. Even if I get into follow-up conversation, I don't know what's the main thing they have in mind or what is something that they need more data on.

@Interactpitch changes it today for all of us founders.

And also, for forever free....

This is quite interesting! Does the AI avatar also sound like the person who is meant to be 'presenting' the deck?

This is pretty cool, but initially I thought the recording was a human narrating the presentation rather than the AI answering additional questions. Is it possible to have both?

Congrats on the launch.
The demo video looks cool.
May I know what tools you used to come up with that video?

Fundraising is stressful enough without guessing what investors actually thought of your deck. Wishing all of best for your launch:)

Hey team, congrats on the launch!

Super curious about the tech under the hood – how customizable is the AI avatar’s voice/scripting, and does it integrate with existing tools like Figma or Google Slides?

Would love to chat more about this product and maybe even test it out for one of our portfolio companies :)

@suyash_karn I love the analytics part of it, not so sure though if i want the avatar/chat in my deck, is it possible to activate just analytics as well?

Turning decks into slide-aware Q&A + real time viewer analytics is slick; at scale the hard part is keeping answers tightly grounded while many investors hit the same deck concurrently.

Best practice: per-slide chunking + cited retrieval (Postgres + pgvector), stream voice/chat via WebRTC/WebSockets, and push engagement events into ClickHouse (or Kafka→warehouse) for fast “time-on-slide + questions” reporting.

Open Q: are you enforcing retrieval strictly to the active slide (with quoted spans/citations) or a global deck KB and how are you approaching SOC2/privacy for investor activity data?

This is a really thoughtful take on fundraising. The idea of understanding how a pitch was experienced, not just sending it out and hoping for the best, feels like a big gap most founders just accept today. The pre-call insight angle is especially interesting knowing where investors spent time or got confused could completely change how the first conversation goes. Curious to see how founders balance this with live pitching, but the positioning as an enhancement (not a replacement) makes a lot of sense.

Congrats on the launch excited to see how this evolves.

Congrats on the launch! How much info do I need to put into the knowledge base to ensure that AI can give satisfying answers for investors' questions? Would AI make something up?

What platform support do you have today? Wondering if VCs can easily view on the go.

Really like the idea of making pitch decks experienced, not just viewed.

Curious how you think about the balance between guidance and agency here:
when an AI avatar explains context, do investors tend to explore more freely – or does it subtly anchor their attention to what you think matters most?

Feels like a powerful tool for reducing ambiguity in early conversations. Congrats on the launch 👏

This doesn’t replace live pitching, and that framing is smart. It makes the pitch deck more explorable and less one-directional. Curious how this evolves with real investor feedback.

👋 Hey Product Hunt!

I’m Huang, the founder & CEO of Typeless.

Since the end of 2025, we’ve launched Typeless on Mac, Windows, and iOS.

In just a few months, tens of thousands of people started using Typeless every day to write messages, emails, notes, long-form content, and for vibe work.

Across all platforms, one request kept coming up again and again:

“When is Typeless coming to Android?”

Today, I’m excited to finally answer that.

Typeless for Android is live 🎉

—

Typeless for Android is the world’s first smart voice keyboard on Android that truly understands you.

You speak naturally. Typeless understands your intent - and turns your raw speech into clear, polished, natural writing in real time.
No commands. No formatting tricks.

It reads as if you carefully wrote it yourself.

When input friction disappears on a phone, everything changes. Your iPhone becomes capable of things that once felt impossible - at 10× the speed.

—

Core features:

• Works everywhere – Any app on your phone: WhatsApp, Slack, Gmail, Notion, Notes, and more

• AI auto-editing – Removes filler words, fixes grammar, and formats perfectly as you speak

• Speak to Edit – Revise, refine, or rephrase by simply saying what you want to change

• Whisper Mode – Dictate quietly and precisely in low-voice environments

• 100+ languages – Speak any language (or mix them); Typeless detects and transcribes automatically

• Personal Dictionary – Names, brands, and custom vocabulary learned once, never missed

• Private by design – Zero data retention. We never store or train on your data. All history stays on your device

—

The keyboard was a mistake.

Typing on a phone is where that mistake becomes painfully obvious.

The on-screen keyboard was a transition.

Voice is the next generation.

This isn’t just a new Android app.

It’s the next step in how people communicate — and create — with their devices.

Welcome to the end of typing.
Welcome to Typeless.

—

I’d love to hear from you:

• Where does Typeless save you the most time on your phone?

• In which moments do you still fall back to typing?

• What’s the one feature you’d like us to build next?

We’re here all day to chat, answer questions, and hear your ideas.

If Typeless makes using your phone feel more effortless, your support would mean a lot to us 💛

- Written with Typeless

Hey Product Hunt!

I'm Krystal from Typeless. Excited to finally be here with our Android launch!

Why we built this:

After launching on iOS, macOS, and Windows, one message flooded our inbox — "I'm on Android. When can I get this?"

Android users were juggling multiple apps: voice recorders, transcription tools, and then spending ages cleaning up the mess in Google Docs.

That workflow is broken.

The core problem:

Mobile typing is painfully slow, and standard voice input just dumps raw text complete with "uh"s, rambling, and zero structure. You’re left fixing everything manually.

Typeless for Android reimagines this completely:

• Works everywhere
Seamlessly available inside any Android app, including WhatsApp, Gmail, Slack, Notion, Notes, and more.

• Intelligent writing, not raw transcription
Typeless understands your intent as you speak. It removes filler words, fixes grammar, organizes your thoughts, and adapts tone in real time.

• Speak to Edit
Want to revise, refine, or rephrase your writing without typing? Just say what you want to change, and it happens like magic.

• 100+ languages with instant detection
Speak any language or switch mid-sentence. No settings to change.

• Personal dictionary
Add your unique words effortlessly, and Typeless won’t miss them.

• Privacy by design 

All processing happens on your device. Zero data collection. Your voice never leaves your phone.

Beta feedback from Android users:

6x faster than thumb-typing for most users

"Finally feels like a native Android experience" - consistent feedback

Top use cases: WhatsApp replies, Gmail composition, capturing ideas in Keep

Android-specific questions for you:

Device optimization – Which Android phone are you using? We want to ensure perfect performance on YOUR device.

Integration depth – Should we prioritize Tasker automation, Bixby Routines, or Google Assistant shortcuts?

Customization – What level of widget customization matters to you? (Size, style, quick actions?)

Ecosystem features – Samsung DeX support? Wear OS companion app? Foldable-specific UI?

Performance tradeoffs – Would you prefer faster processing with the internet, or slightly slower but fully offline?

I'm here all day discussing technical details, feature priorities, and Android-specific optimizations. Tell me how
Typeless fits into your Android workflow!

Download: https://www.typeless.com/

Krystal

Dictated with Typeless for Android:>

Works perfectly on my Pixel 8. The AI removes all my filler words automatically and makes me sound way more articulate than I actually am.

👋Hey Product Hunt!

I'm Stellar from the Typeless marketing team, and today is a day we've been waiting for.

Here's what happened:

After launching on iOS, MacOS, and Windows, our inbox became filled with messages from Android users. Not just "when's Android coming?"—but stories about why they needed it.

A product manager who brainstorms features during his bike commute. A journalist who drafts articles on the subway. A teacher who responds to parent emails between classes, phone in one hand, coffee in the other.

They all said variations of the same thing: "I chose Android for freedom and flexibility. Give me that same freedom from typing."

So we built Typeless for Android. And the experience? It's exactly what Android deserves. Your messy thoughts become ready-to-send messages. No editing. No cleanup. Just speak, and it's done—polished, clear, and sounds exactly like you carefully typed it.

The moment that convinced us:

Our beta tester Maria shared: "I was grocery shopping when my boss messaged about an urgent client issue. Cart full, hands occupied. I just... spoke into my phone. A complete, professional response appeared in seconds—no 'um's, no rambling, perfectly formatted. I hit send without changing a single word and continued shopping. That's when it hit me—this isn't just convenient. It's liberating."

That's the difference. Other voice tools transcribe what you say. Typeless transforms it into what you mean to send.

That's what we're launching today.🎉🎉

What we want to know:

• How many ideas have you lost because typing on your phone felt like too much effort?

• What would shift in your daily routine if communication was as natural as speaking?

• Do you ever ignore messages simply because typing a proper response feels exhausting?

We're here all day to listen, share stories, and answer everything. If you've ever felt your thoughts racing ahead of your thumbs—if you've chosen Android because you value doing things your way—your support today would mean the world to our team.

Experience it yourself: Typeless for Andorid

Stellar

Dictated with Typeless for Android

Finally you guys release Typeless for Android after a long wait! savor for android user like me! any difference in funcs?

Wow, finally. We have typeless on Android yay.

Definitely going to give this a try for at least notes and drafting.

Is there a way to adjust how much the AI alters what I say? How heavy the editing is?

Also, what runs locally on my device and what is done in cloud? and for whatever runs locally, do you have minimum requirements listed?

Just dictated a whole email to a client while walking my dog. Sent it without editing. This is genuinely the future of mobile productivity.

Works perfectly on my Pixel 8. The AI removes all my filler words automatically and makes me sound way more articulate than I actually am.

The Quick Settings integration is so much more convenient than I expected. I use it constantly now instead of opening the full app.

I’m a big fan of Typeless. It lets me accurately input what I want to say even when typing isn’t convenient, like when I’m walking.

This is exactly what Android needed. I can finally dictate messages while walking and they actually sound professional, not like rambling nonsense.

Huge congrats on the launch! The voice recognition works even when I'm speaking quietly in the library. Don't have to raise my voice or repeat myself multiple times.

Congratulations on the launch! I’m looking forward to following all your updates.

It would be even more powerful if it could attach a place or address directly from Google Maps when dictating - super useful for messages and emails.

I'm an iphone user and may just not know - but how is this different than wisprflow?

I’ve been waiting for something like this on Android for ages. Most voice typing feels clunky. This actually sounds… human.

Shipping a voice keyboard IME that “cleans up” speech in real time gets gnarly at scale: ultra-low-latency streaming + consistent tone while thousands of users dictate across wildly different apps.

Best-practice: VAD + streaming ASR (on-device whisper.cpp/NNAPI/Core ML where possible) with partial hypotheses, then a fast rewrite pass; add automated quality/latency evals + guardrails to keep the “zero retention” trust story intact.

Open Q: are you doing on-device-first ASR with optional cloud rewrite, and how do you avoid app context leakage while still personalizing style?

Congrats! I've been waiting for an app like Typeless 2 for ages! Finally, something that makes voice input feel effortless without the usual hassles.
How well does it handle non-English languages or accents?

Wow... the first Evernote launch since 2020...! And since Bending Spoons acquired them.

Bold — drenched in AI!

Glad I exported my notes years ago, looks like they're now being held hostage! 🫠

Thanks for posting this @chrismessina . Funny thing is, if you browse around a bit you can still find the old UI from years ago. These are the settings and profile pages, for example:

NEW

https://accounts.evernote.com/settings/account/summary

OLD

https://www.evernote.com/Settings.action

I wonder how come they never removed the old batch of pages from the server... 🤔

I any case, I am over the 50 pages limit for the free version... 😔 - CIAO!

Software that was great, that you continue to make worse.

Big moment for Evernote after a long pause – congrats on the v11 launch.

Curious how you’re thinking about when AI should step in vs. when notes should stay quiet. In my experience, a lot of thinking happens before ideas are ready to be summarized, searched, or "understood."

How do you balance augmentation with not pulling meaning out of notes too early?

Got to use Evernote a lot back in the days, why should I back to it? Congrats on the launch anyway :)

Ah.. Evernote! It's always cool when "older" brands get a new life with new owners. I'll check it out, if only to see my 2011 notes first.

Hey Product Hunt 👋
Ron here, the creator of Dopamine.chat.

All you want is to “clean your inbox”, “summarize customer feedback every morning,” or “create and post daily content”.

But every platform promises this, and they all end up feeling the same.
You sign up, get excited… and then it becomes insanely hard to actually get the work done.
Too many steps. Too many settings. Too much overhead.
So the task gets postponed… waiting for the next platform that might solve it.

I felt this pain too, deeply.
That’s why I built Dopamine.
First for me. Then to more people who felt the same frustration.
Dopamine is optimized for one thing: giving you value fast.
No steep learning curve. No complex setup. No “build a full automation system before you get a result.”
Just simple, powerful agents that start helping you from day one.

Here’s what Dopamine lets you do:
✅ Build mission-specific agents in few clicks
✅ Connect them to your real data with built-in integrations
✅ Customize how each agent thinks and behaves
✅ Combine agents into workflows or multi-agent teams
✅ Share your agents with teammates or communities

What people are already building:
1️⃣Research agents that summarize, compare, and digest info instantly
2️⃣Admin & ops automations that clear repetitive tasks
3️⃣Content agents for drafting, repurposing, and brainstorming
4️⃣Personal assistants for planning and organizing
5️⃣Startup workflows for onboarding, support routing, CRM tasks, and more

I’d genuinely love to hear what you built, what surprised you, and what still feels missing.
Can't wait to see what you do with dopamine.chat 🙏

- Ron

Give me a way to auth and give read/write access to fly.io, railway, supabase, one or all Github repos, Stripe and a few other things and enpowr an agent to debug. Do this and become a billionnaire this year.

Great name! Honestly, getting an automation to finally work IS a hit of dopamine:D Most tools feel like homework, but this looks like it’s actually fun to use. Which of the pre-built agents is the most popular for beginners to start with?

Congrats on the launch! The promise of 'value fast' is great, but simple tools often hit a ceiling quickly. How does Dopamine handle complex, multi-branch logic without eventually turning into the very 'heavy setup' platform you're solving against?

Love your brand name 😄 Do you have integrations to use agents outside of your platform? For example: implementing an agent inside of a CMS?

The fact that you built this to solve your own frustration shows in the simplicity. That usually translates well to users.

This feels like one of those products where the real magic shows up after a week of use. Any unexpected learnings from early users so far?

Simply valuable. In a world where every second new product is an AI agent, dopamine chat managed to actually make automations of agents easy. I use it for my own product with my emails ingestion.

I’m waiting to see what will be further released

Very useful product
1 recommendation: For me i would hgave been better if it also told source of its research in reddit , X and web search

Hey Product Hunt 👋

We’re excited to introduce Figy AI, a new learning tool from the MyLens AI team.

One of the biggest challenges in education is that learning needs both great content and real engagement. Having the right information isn’t enough—if learners aren’t engaged, it’s hard to truly understand and remember it over the long term.

Figy helps bridge that gap by turning any content—topics, notes, articles, or documents—into visual, AI-powered flashcards that are easy to explore, expand, and designed to help learning stick.

What makes Figy really different is that the flashcards don’t stop. You can keep going deeper into a topic for as long as you want, while the AI continues to challenge you with new questions and updated cards—helping you truly understand, not just memorize.

We built Figy AI to make learning clearer, more engaging, and something people actually want to come back to.

Thanks for checking out Figy 💜

Just tried Figgy AI out - it is so easy and yet powerful way to create activities for educators, I wish I had it when I used to be a teacher 😻

this is called gold mate! keep making such products like these

The concept of flashcards that 'keep going deeper' is a smart differentiator from static AI generators like NotebookLM. How does the AI decide when to push a user to a more advanced sub-topic versus just reinforcing the basics?

Congrats on the launch. AI flash cards are fun. What’s your fav use case among your customers?

@ardalan2 Congratulations. And happy product launch.

As an early engineer at Rippletide, I've spent countless hours testing AI agents and getting frustrated with all the vague performance metrics.

That's why we built Rippletide CLI: a terminal tool that lets you benchmark your AI agent directly from the command line. It generates questions from the agent's own knowledge, supports reproducible test sets, and gives clear KPIs on hallucinations.

Everything runs in real time with automatic evaluation and detailed reports, so you actually see where your agent struggles.

Would love to hear what the PH community thinks and get feedback from fellow AI builders 🚀

Hi all, very excited to present an Agent evaluation module today!

As AI engineers, my team and I struggled to reliably tell whether the latest version of an agent was actually performing well or not.
So we built a module to evaluate agents, and we’re open-sourcing the hallucination measurement part.

How it works:

1 – Connect your agent
Use our CLI to provide your agent endpoint (localhost works).
Connect the data your agent needs. Today, we support PostgreSQL databases, internal APIs, and Pinecone as a vector store. If you’d like to add a new source, feel free to open a PR on the repo.

2 – Launch tests
Tests are automatically generated to evaluate your agent’s behavior and make sure no agent possible wrong behavior nothing is left out - stay safe. You can also add your own test set if needed.

3 – Understand what failed
For each test question, we check every fact in the agent’s answer and verify whether it has a reference in our graph. We then explain where additional data is needed to improve your agent.

You can then improve the agent on the Rippletide platform and re-test it. We believe that when an agent reaches less than 1% hallucinations, it can be deployed in production. Some use cases require 0.1% or even 0.01%, depending on volume or industry.

Feel free to ask any questions, or reach out if you’d like to know more about what we’re building.

Cheers,
Yann

Congratulations on the launch 🚀

does the benchmarking feature allow us to compare historical runs side-by-side to track drift over time?

Congrats on the launch! A lof of the times, the evaluation metrics are heavily related to the use cases, will the generated questions and evals be specially tuned for specific scenarios/ industries?

Nice tool to get awar from most of noises ! how frequent will trace refresh the content feed?

Congrats on the launch! Would love to see deeper and integrated info instead of original content :)

Nice. How big is the memory for storing the info

We open a new tab dozens of times every day. We built this extension to make those moments a little more beautiful and inspiring, and hopefully help us think and build better.

It showcases evocative visual thinking, a curated collection of paintings, and tweet-style quotes from history’s brightest minds.

Some details

• More than 200 works and growing, with 80% devoted to ideas and quotations and 20% to timeless art.

• Every new tab fades in softly, adding a refined touch to the experience.

• You’ll receive new content without needing to update.

• Click anywhere on the page to see new pieces, like advancing slides in a presentation.

We’ve open-sourced the code to create your own extension.
https://github.com/oscarotero/visual-tabs

Would you have any ideas to improve it or suggest additional content formats?

Thanks!

congrats on the launch! I love this idea, feels very useful and meaningful. motivational quote or a piece of art could be really helpful midwork. great, I wish you all the best with it!

Love that it’s minimal and focused on inspiration without the usual noise of bookmarks or widgets. Perfect for keeping that creative spark alive between client calls and code reviews ahahah

Quick question: How often does the artwork/quote rotate? Is it daily, per new tab, or customizable?

Hey, it's Matt from MCPJam 👋.

We started MCPJam because the dev experience around building ChatGPT apps, MCP apps, and MCP servers wasn't great.

To develop a ChatGPT app, you had to get a ChatGPT pro subscription, ngrok you local MCP server, then connect it remotely on ChatGPT. There were also no good ways to test MCP servers and see how they behave in a real production environment.

That inspired us to build our open source inspector at MCPJam. We introduced:

• Local ChatGPT and MCP apps emulator.

• LLM playground to test your MCP server / apps in a real chat environment with any LLM

• OAuth debugger to visualize MCP authorization at every step

and a bunch of other features that addresses the pains of MCP development.

If you're building a ChatGPT app, MCP app, or MCP server, I highly invite you to give us a try! Feel free to email me if you have any questions or feedback.

GitHub: https://github.com/MCPJam/inspector
Email: matthew@mcpjam.com

Our team at Tadata builds MCP servers every day. We maintain one of the leading open source MCP projects and are shipping a large number of MCP servers for our agent builder. And we're huge fans of the OG MCPJam tool for debugging OAuth issues -- the visual flow is so helpful! Excited to try out the new stuff.

MCPJam is an amazing product with polished UX that has become our daily driver for troubleshooting. I only wish we had the OAuth inspector sooner, it would have saved us days of frustration! We now trust it as the de facto standard for validating new features. This is a must-have tool that feels like the start of an 'Android Studio' for ChatGPT apps. Good luck

Congratulations on the launch!

Electric! Great work, fellas!

Let's go!! Love our community and what we've built.

woop woop

Nice focus on the local development workflow - the "no more ngrok or ChatGPT subscription needed" angle is a real pain point.

The tool inspector for resources, prompts, and OAuth flows looks especially useful. Debugging OAuth in MCP integrations has been frustrating in my experience.

One question: does the LLM playground support streaming responses, or is it request/response only? We've found streaming adds complexity when testing tool interactions.

Great launch - the MCP tooling ecosystem needs more developer-focused products like this.

Congratulations on the launch! MCPJam was clutch when we were building and testing MCP apps at Noodle Seed.

How many languages are supported as of now?

Congrats!!

Quick question, does it support uploading existing local video files (not just YouTube downloads)?

Would love to chat more about it and potentially try it out for some upcoming multilingual projects we have :)

Two-pass translation, local processing, BYO keys, and open source is a rare combo that actually respects developers and creators.

Pricing feels fair and approachable. Excited to see where this goes, Kudos on the launch

Congrats on the launch, Mikey.

The two-pass translation idea makes a lot of sense especially for video where literal translations usually sound off.Nice touch treating it as translation + editing instead of just running text through a model once.Also like the pay-for-usage + BYO keys approach. Feels fair and flexible compared to most tools in this space.

Slang and casual phrasing usually break most tools interested to see how this handles that.

Nice work shipping this.

the GPT + Claude combo for quality checks is a smart way to reduce translation hallucinations. do users need to plug in their own API keys to handle the volume or is that managed on your end?

Are there options for cuter icons like cats/ dogs?

The beauty of this idea is in the simplicity. Which I think is in line with the whole idea of MacOS. Nice work!

finally, I can grab a coffee without worrying that my AI agent is deleting files back at home. can I hit a kill switch instantly from my phone if I see it going off the rails?

Looks useful! Does this handle multiple repositories and multiple checkouts of a single repository? I'm assuming so from reading the landing page. How do you use this, do you just make sure your laptop is on all the time? Could we deploy to a remote server?

Hey PH! 👋
The indie dev paradox: You finally have AI writing code for you. Revolutionary. Life-changing. And yet… you're still chained to your laptop like it's 2019.
Your agent finishes at 2am. You find out at 9am. Three PRs are blocked. Your momentum is dead. Very build-in-public of you.
So I built relayd — a tiny daemon that connects your local Codex to a mobile dashboard.

🔔 Push notifications when agents need you
📱 Mobile-first PWA (no app store politics)
👀 Full context — diffs, errors, existential dread
⚡ One-tap continue — like a tamagotchi that writes code

Pricing that makes sense:

Free: 1 relay, 1 workspace, unlimited threads
Pro ($5/mo): Unlimited everything. Less than your cold brew.

Built by an indie dev, for indie devs. No enterprise bloat.
Would love your feedback!

Congrats! The freedom sounds nice, but I usually need a full monitor to properly audit what my agent wrote. How granular is the mobile diff view? I'd be worried about 'approving' complex changes from a phone screen.

Shared this internally with the team at Crisp as we're all big on coffee 😂

Sent it to my friend who admires coffee :D

ohhhhh this one is such a good idea ahahah

Does it let you log custom caffeine sources (like specific teas or chocolates) with adjustable mg amounts? And how does the sleep analysis tie into the caffeine data – any cool visualizations for correlations?

Congrats on the launch! This is an interesting idea. I think coffee no longer works for me, but I am curious again to start drinking coffee and use this app :)

Modeling caffeine half-life decay on Apple Watch + Apple Health integration is a super practical way to connect “what I drank” to “why I slept poorly.” The scale pain is accuracy (different drinks/servings) + adherence; best practice is quick-add presets + barcode/shortcut intake logging, and use HealthKit sleep signals to auto-tune a user’s half-life model over time. Any plans to ship an Android/Wear OS version? I’ll be looking for it.

Love the half-life decay approach here — it’s way more useful than just logging “cups.”

Is the caffeine half-life a fixed value, or can it be personalized (weight/sleep sensitivity/metabolism)? Customization would be huge.

Do you integrate with rings and google health?

interesting one for all coffee lovers :)

Congtarulations for your launch!

Great app for white-collar professionals ☕️

Local-first “build the app by conversation” + explicit filesystem approvals is a strong take on personal software. The scale pain is safety + determinism as users iterate (prompt drift → broken apps), and best practice is: a typed app schema + versioned diffs/migrations, plus a permissioned tool layer (capabilities) and replayable build logs for debugging. How are you representing apps under the hood (React/Tauri/Electron + JSON schema?), and will you add one-click export/share of an app as a signed bundle?

Hey there! Congrats on this launch :)

Does it have any smart reminders (time/location-based) or streak recovery features for when life inevitably derails a day or two?

Congrats on the launch! The habit tracking space is incredibly crowded. What is the one specific feature or philosophy in Habitz that convinces users to switch from their current tools?

Hey Product Hunt! 👋

I built this because I use Google Drive for everything and there is no native preview for 3D files. The few apps that exists are just filled with ads/broken or look like they were made for Microsoft XP.

My goal was to build the "Google Docs for 3D Files" a clean, fast way to preview OBJ, STL, GLB and tons of other formats files directly inside Drive and Gmail without downloading them.

Some features :

AI Rendering: You can now take a snapshot of your mesh and instantly generate a photorealistic render using AI.

Collaboration: Added "Google Docs-style" commenting and measurement tools directly on the model.

I’m a solo dev and would love to hear your thoughts. Does the viewer feel fast enough for your workflow?

This solves the 'I can't open this file' email from clients. Does the viewer render smoothly within the mobile Gmail/Drive apps? That would be a game-changer for quick approvals on the go.

nice! very neat missing drive feature. Any plans for share links (view/comment) plus @mentions?

Nahhh can't believe it! It's really cool man! 3D in google drive blowed my mind. Just guessing what's the business model behind?

As a 3D company i tried it and used it, the interface is really clean and It’s a breath of fresh air to see such an app in the 3D sector. The renderer with AI is also really useful, it help really to connect the idea of AI and 3D point of view which is not so evident. I tried with my colleagues the commenting options and i find it really usefull, i can for example show where i find a problem on a mesh, or which part need a modification without making a screenshot and drawing on it hahaha. Finally the whole drive and gmail interface is really usefull ! Sorry for my bad english i'm french ^^

Today I’m launching Stroke 🎉
It’s built to be simple, fast, and focused on real-world use cases.
Your feedback will help shape what comes next 🚀

Help me shape it better. Yes it's open source
Github : https://github.com/abhix4/stroke

Manually animating SVG paths is usually a nightmare. Does this export pure CSS/SVG code that I can just copy-paste, or does it require a JS runtime to play the animation on my site?

simple and elegant, great work! would it support picture upload later?

It's cool, but the UI could do with a tweak. Namely, I'd loop the animation or allow you to run it again. On Desktop the results were not visible the first time I used it.

This hits a real parent pain point.

What standards/databases do your risk flags come from? And how often do you update info when brands change formulas across countries?

the focus on hidden sugars and additives is crucial, especially with how confusing packaging can be these days. I’m curious about your ingredient criteria. are you leaning on strict EU standards or a custom pediatric framework to flag these risks?

Great idea. I think that in Europe we have sort of strict rules when it comes to food, but definitely good for us as you cannot be 100% sure what everything the food contains. Nice one!

Hey Product Hunt 👋

Excited to share Raycaster with the community today 🚀

I’ve been closely involved with this launch and had the chance to see the product evolve from early internal workflows to what it is now. What really stood out to me is how Raycaster treats regulatory and drug-development documents like a true system, not just isolated files.

Instead of one-off AI drafting, it works at a project level across hundreds of PDFs, Word, Excel, and PowerPoint files, keeping dependencies, citations, and versions in sync. In industries where a single inconsistency can delay approvals or trigger rework, this approach genuinely matters.

From diff-aware edits to traceable citations and version control that actually makes sense for regulatory teams, Raycaster feels purpose-built for high-stakes document workflows rather than a generic AI wrapper.

Huge credit to the team for tackling a problem most people underestimate: the documents are often the bottleneck, not the science.

Would love to hear feedback, questions, or real-world use cases from the PH community 🙌

Happy to help clarify anything in the comments!